ERROR - Please make sure that you have provided the correct assembler type and the correct path to the binning result file in the correct format. about graphbin

[jolespin]

Thanks for including the prepResult.py to preprocess everything.
I'll give it a try. So it truncates all of the contig identifiers?

In the future would it be possible to make this a bit more seamless to integrate into existing pipelines? For example, I'm running metabat2 and maxbin2 then I feed everything into DAS_Tool which takes in [scaffold]\t[bin] tsv files with the identifiers unmodified (created using https://github.com/cmks/DAS_Tool/blob/master/src/Scaffolds2Bin_to_Fasta.sh).

If graphbin could take in unmodified identifiers and do the conversion in the backend automatically it would make it much more practical to use in a complicated pipeline where similar versions to the required files are already generated. In an ideal scenario, it would be awesome to have DAS_Tool take in the pre/post graphbin binning for it to calculate a consensus for everything w/ as few steps as possible.

[Vini2]

Dear @jolespin,

Thank you for posting the issue with the relevant data. I went through your files and found that the binning results file scaffolds_to_bins.csv is not in the format required as input for GraphBin. As the documentation on input format outlines, the binning output file should have comma separated values (contig_identifier, bin_number) for each contig. contig_identifier should follow the pattern specified and bin_number should be a number without any letters or other characters.

GraphBin provides a support script named prepresult.py to generate similar files once the initial binning output folder is provided. You can refer to support/README.md for more details.

Let me know if this won't solve the issue.

I also fixed the --assembler argument and now you can provide SPAdes as well.

Thank you!

[jolespin]

Thanks for getting back to me. I just tried it with the bins relabeled and got the same error.

(graphbin_env) jespinozlt2-osx:GraphBin jespinoz$ head /Users/jespinoz/Downloads/files/scaffolds_to_bins.relabeled.csv
NODE_14_length_9160_cov_4.404833,0
NODE_15_length_7837_cov_3.578386,0
NODE_16_length_7509_cov_4.024014,0
NODE_17_length_7205_cov_3.142238,0
NODE_18_length_6867_cov_2.992073,0
NODE_19_length_6076_cov_3.252782,0
NODE_20_length_5936_cov_3.975514,0
NODE_21_length_5583_cov_3.496020,0
NODE_22_length_5539_cov_3.360686,0
NODE_24_length_5345_cov_3.093384,0

Most of the binning programs output alphanumeric strings so it seems a bit more difficult to integrate this into existing pipelines. Does prepresult.py format the graph and contigs/scaffolds as well?

[Vini2]

As the documentation on input format outlines, the (contig_identifier, bin_number) pairs should be in the form

NODE_1,1
NODE_2,1
NODE_3,1
NODE_4,2
NODE_5,2
...

Note that the numbering of bins should start from 1

prepresult.py handles all these cases and outputs the binning result as shown. You can first run the prepresult.py script as

python prepResult.py --binned /path/to/folder_with_binning_result --assembler spades --output /path/to/output_folder

Since MaxBin2 outputs the binning result to .fasta files for each bin, you can provide the path containing these .fasta files to the --binned option.

After getting the formatted binning result, you can input it to GraphBin.

Let me know if it still doesn't work.

[Vini2]

That would be a great option. I will modify GraphBin to support a more generic convention for inputs and outputs as the one you mentioned [scaffold]\t[bin] where scaffolds/contigs and the bins will have their original identifiers and add a new version of GraphBin.

Thank you very much for the suggestion! I will keep this issue open until I fix it. In the meantime, please let me know if it still doesn't work.

Thank you!

[JSSaini]

Hello, I have a similar issue. I can see that it is reporting my binning file (file attached) to contain two bins . However, I can confirm that my binning file has just one bin number. Please see the file attached.

clustering_merged_mod.zip

2021-04-14 17:44:53,372 - INFO - Welcome to GraphBin: Refined Binning of Metagenomic Contigs using Assembly Graphs.
2021-04-14 17:44:53,375 - INFO - This version of GraphBin makes use of the assembly graph produced by SPAdes which is based on the de Bruijn graph approach.
2021-04-14 17:44:53,376 - INFO - Input arguments:
2021-04-14 17:44:53,377 - INFO - Assembly graph file: /home/users/s/saini7/scratch/MS2/Anvio2/Spades_13_15p5_Map_GBII/assembly_graph_after_simplification.gfa
2021-04-14 17:44:53,378 - INFO - Contig paths file: /home/users/s/saini7/scratch/MS2/Anvio2/Spades_13_15p5_Map_GBII/contigs.paths
2021-04-14 17:44:53,379 - INFO - Existing binning output file: /home/users/s/saini7/scratch/MS2/Anvio2/conc2/spa_concoct2/clustering_merged_mod.csv
2021-04-14 17:44:53,381 - INFO - Final binning output file: ./
2021-04-14 17:44:53,382 - INFO - Maximum number of iterations: 100
2021-04-14 17:44:53,383 - INFO - Difference threshold: 0.1
2021-04-14 17:44:53,384 - INFO - GraphBin started
2021-04-14 17:44:53,393 - INFO - Number of bins available in the initial binning result: 2
2021-04-14 17:44:53,395 - INFO - Constructing the assembly graph
2021-04-14 17:44:53,700 - INFO - Total number of contigs available: 2655
2021-04-14 17:44:57,181 - INFO - Total number of edges in the assembly graph: 10554
2021-04-14 17:44:57,184 - INFO - Obtaining the initial binning result
2021-04-14 17:44:57,188 - ERROR - Please make sure that you have provided the correct assembler type and the correct path to the binning result file in the correct format.
2021-04-14 17:44:57,188 - INFO - Exiting GraphBin... Bye...!```

[Vini2]

Hello @JSSaini,

Your binning file has a header contig_id,cluster_id which should not be included. This is why GraphBin counts cluster_id and 43 as 2 bins. Also I noticed that your contig ids are in the form c_00000000000*. These contig ids should be the form NODE_* as found ids found in your contigs.paths file. You can double check that as well.

GraphBin provides a support script named prepresult.py to generate the binning file once the initial binning output folder is provided. You can refer to support/README.md for more details.

Let me know if this won't solve the issue.

Thank you!

[JSSaini]

Still the same error, this time I made sure I have given the appropriate binning file. Have you tried this tool with concoct binning results? Thank you.
conc_graph.txt

2021-04-16 13:27:36,494 - INFO - Welcome to GraphBin: Refined Binning of Metagenomic Contigs using Assembly Graphs.
2021-04-16 13:27:36,591 - INFO - This version of GraphBin makes use of the assembly graph produced by SPAdes which is based on the de Bruijn graph approach.
2021-04-16 13:27:36,593 - INFO - Input arguments:
2021-04-16 13:27:36,595 - INFO - Assembly graph file: /home/users/s/saini7/scratch/MS2/Anvio2/Spades_13_15mw_Map_GBII/assembly_graph_with_scaffolds.gfa
2021-04-16 13:27:36,596 - INFO - Contig paths file: /home/users/s/saini7/scratch/MS2/Anvio2/Spades_13_15mw_Map_GBII/contigs.paths
2021-04-16 13:27:36,597 - INFO - Existing binning output file: /home/users/s/saini7/scratch/MS2/Anvio2/conc2/spa_concoct2/conc_graph.csv
2021-04-16 13:27:36,599 - INFO - Final binning output file: ./genomes/
2021-04-16 13:27:36,600 - INFO - Maximum number of iterations: 100
2021-04-16 13:27:36,602 - INFO - Difference threshold: 0.1
2021-04-16 13:27:36,603 - INFO - GraphBin started
2021-04-16 13:27:36,609 - INFO - Number of bins available in the initial binning result: 1
2021-04-16 13:27:36,610 - INFO - Constructing the assembly graph
2021-04-16 13:27:36,852 - INFO - Total number of contigs available: 2845
2021-04-16 13:27:37,584 - INFO - Total number of edges in the assembly graph: 11581
2021-04-16 13:27:37,585 - INFO - Obtaining the initial binning result
2021-04-16 13:27:37,589 - ERROR - Please make sure that you have provided the correct assembler type and the correct path to the binning result file in the correct format.
2021-04-16 13:27:37,590 - INFO - Exiting GraphBin... Bye...!```

[Vini2]

Hello @JSSaini,

Sorry for getting back to you late. I fixed some errors in the indexing of the input bins. Can you please get a git pull of the latest code of GraphBin and check your results? It should work fine now.

Let me know if this won't solve the issue.

Thank you!

PS: I haven't added a new release yet with the fix as I plan to add some more fixes. I will add a new release including all the fixes ASAP.

[Vini2]

Hello @jolespin and @JSSaini,

I have added a new release of GraphBin with fixes to this issue. Please give it a try when you have time.

https://github.com/Vini2/GraphBin/releases/tag/v1.4

Thank you very much for the input.

[jolespin]

"Fixed naming of SPAdes contigs to have original contig identifiers and updated prepResult.py to reflect this fix." 🙏🏽 thank you! This will make it much easier to implement into existing pipelines

[Vini2]

Closing issue after fixing.